AnomalyXL: Tiny Recursive Language Model Hits SoTA on Long-Context Time-Series Anomaly Detection
samsja19 · x · 2026-10-09
A new paper shows frontier time-series LLMs struggle with precise anomaly localization in long contexts. The authors trained a tiny Recursive Language Model to SoTA using AnomalyXL, an infinitely scalable synthetic time-series generator with programmatic ground truth, and the approach transfers to real-world anomaly detection in software observability, ECG, and sleep data.
More from Research
- Agent Plasticity paper shows Claude learning reusable rules from NetHack and Chess failures — anirudhg9119 · 2026-10-09
- Leak hints OpenAI runs multi-agent swarms on time budgets, not token budgets, and treats it as IP — maksym_andr · 2026-10-09
- 100+ Mathematicians React: OpenAI's Quasi-Riemann Result "Almost Unbelievable" — littmath · 2026-10-09
- Epoch launches Automation Reports: Claude Fable 5.1 and GPT-6 Astra lead but can't automate its research — scaling01 · 2026-10-09
- Iris-3B: pixel-space diffusion offers no edge over latent models, study finds — speridlabs · 2026-10-09
- Meta's MIMESIS: 9B user simulator beats GPT-5.5 for training interactive agents — facebook · 2026-10-09